Parent–Adolescent Discrepancies in Reports of Parenting and Adolescent Outcomes in Mexican Immigrant Families

Abstract

Parents and adolescents often have discrepant views of parenting which pose challenges for researchers regarding how to deal with information from multiple informants. Although recent studies indicate that parent–adolescent discrepancies in reports of parenting can be useful in predicting adolescent outcomes, their findings are mixed regarding whether discrepancies relate to more positive or more negative adolescent outcomes. This study examined the longitudinal implications of parent–adolescent discrepancies in reports of parenting (warmth, monitoring, and reasoning) on adolescent behavioral, psychological, academic, and physical health outcomes among Mexican immigrant families in the United States. Participants were 604 adolescents (54% female, M_age.wave1 = 12.41 years) and their parents. Taking a person-centered approach, this study identified distinct patterns of parent–adolescent discrepancies in parenting and their different associations with later adolescent outcomes. Adolescents’ more negative perceptions of parenting relative to parents were associated with more negative adolescent outcomes, whereas adolescents’ more positive perceptions relative to parents related to more positive adolescent outcomes. There were also variations in discrepancy patterns and their associations with adolescent outcomes between mother–adolescent vs. father-adolescent dyads. Findings of the current study highlight individual variations of discrepancies among parent–adolescent dyads and the importance of considering both the magnitude and direction of discrepancies regarding their associations with adolescent well-being.

Introduction

In clinical and developmental studies of children and adolescents, a strong research design often involves the use of multiple informants; however, the reports of multiple informants (e.g., parents, children) on the same construct are typically discrepant with only modest correlations (Achenbach et al. 1987; De Los Reyes et al. 2015; Korelitz and Garber 2016; Taber 2010). Such informant discrepancy poses challenges for researchers and clinicians because it could lead to very different research conclusions and clinical decisions (Barker et al. 2007; Kim et al. 2013; Padilla-Walker et al. 2012; Taber 2010). Prior studies specifically investigating informant discrepancy primarily focus on discrepant reports of child mental health problems (Achenbach et al. 1987; De Los Reyes et al. 2015). However, the issue of informant discrepancy is also critical in parents’ and adolescents’ reports of parenting which has gained increasing attention in recent research on parenting (De Los Reyes and Ohannessian 2016; Korelitz and Garber 2016; Rescorla 2016; Taber 2010).

Although many early studies tend to dismiss informant discrepancies as merely measurement error, an increasing number of recent studies indicate the utility of parent–adolescent discrepancies in reports of parenting for predicting adolescent outcomes (Abar et al. 2015; De Los Reyes and Ohannessian 2016; Gaylord et al. 2003; Guion et al. 2009; Maurizi et al. 2012). Nevertheless, in extant literature, there is an extensive debate regarding whether parent–adolescent discrepancies in parenting are adaptive or maladaptive for adolescent developmental outcomes, that is, whether discrepancies relate to more positive or more negative adolescent outcomes. Thus, the current study seeks to shed light on this debate by investigating two specific research aims. First, we take a person-centered approach to explore distinct groups of mother–adolescent and father–adolescent dyads with different patterns of discrepancies in parenting that span three key parenting dimensions: warmth, monitoring, and inductive reasoning. Our second aim is to comprehensively examine how different discrepancy patterns relate to adolescent behavioral, psychological, academic, and physical health outcomes. Although scholars have suggested that the extent of and implications of parent–adolescent discrepancies may vary across ethnicity and socioeconomic status (SES), there is a paucity of studies on low-SES Mexican American population. The current study thus focuses on low-SES Mexican immigrant families, as Mexicans represent the largest immigrant population in the United States (U.S. Census Bureau 2010).

Parent–Adolescent Discrepancies in Parenting and Adolescent Outcomes

Based on extant inconsistent findings regarding the implications of parent–adolescent discrepancies in reports of parenting on adolescent outcomes, the modified Operations Triad Model proposes that parent–adolescent discrepancies in parenting can be adaptive or maladaptive for adolescent outcomes (De Los Reyes and Ohannessian 2016). The adaptive hypothesis proposes that parent–adolescent discrepancies in parenting may be adaptive for adolescent development because discrepancies may indicate a normative developmental process (De Los Reyes and Ohannessian 2016; Leung and Shek 2014; Sher-Censor et al. 2012). During adolescence, there is an increasing need for adolescents to achieve autonomy and independence, and thus parent–adolescent relationships undergo renegotiation (Smetana et al. 2006). Meanwhile, adolescents develop more advanced cognitive abilities, which allow them to take and question others’ perspectives. Thus, adolescents may diverge from their parents’ views about parenting practices and family relationships as a way to assert their independence and autonomy (Guion et al. 2009; Leung and Shek 2014). Consequently, as a normative developmental process, parent–adolescent discrepancies may relate to more positive adolescent outcomes. Consistent with this adaptive perspective, prior studies found that greater mother–daughter discrepancies in perceptions of maternal parenting behaviors were correlated with higher levels of daughters’ individuality (Sher-Censor et al. 2012), and father–son discrepancies in their reports of family functioning were linked to higher levels of male adolescents’ self-competence (Ohannessian et al. 2000).

The maladaptive hypothesis, in contrast, suggests that parent–adolescent discrepancies in parenting may be maladaptive for adolescent development because discrepancies may reflect problems in family functioning (De Los Reyes and Ohannessian 2016; Ehrlich et al. 2015; Maurizi et al. 2012). That is, discrepancies may be due to low parent–adolescent relationship quality characterized by high levels of alienation and conflict or a lack of communication between parents and adolescents (Ehrlich et al. 2015; Maurizi et al. 2012). When parent–adolescent relationship quality is low, adolescents may be less understanding of parental behaviors and thus perceive parenting practices less positively relative to their parents. Therefore, as an indicator of family functioning problems, parent–adolescent discrepancies in parenting may relate to more negative adolescent outcomes. Most prior studies are supportive of this maladaptive perspective by demonstrating a significant association between parent–child discrepancies in parenting and more negative adolescent outcomes (Abar et al. 2015; Córdova et al. 2014, 2016; Guion et al. 2009). For example, two studies involving Latino families demonstrated that parent–adolescent discrepancies in positive family functioning measures (e.g., positive parenting) were related to more adolescent reported HIV risk behaviors such as drug and alcohol use and risky sexual behaviors (Córdova et al. 2014, 2016).

According to the modified Operations Triad Model (De Los Reyes and Ohannessian 2016), both adaptive and maladaptive manifestations of discrepancies may exist depending on various conditions, for example, discrepancies in some parent–adolescent dyads may be adaptive whereas discrepancies in other parent–adolescent dyads may be maladaptive for adolescent development. The current study aims to identify such conditions, thus addressing several limitations in prior studies of parent–adolescent discrepancies.

Limitations in Prior Studies of Parenting Discrepancies and Adolescent Outcomes

The mixed findings about the implications of parent–adolescent discrepancies in parenting for adolescent outcomes may be due to several limitations in prior studies. First, most prior studies take a variable-centered approach to examine how difference scores of parents’ and adolescents’ reports of parenting relate to adolescent outcomes (Córdova et al. 2014; Gaylord et al. 2003; Guion et al. 2009; Reidler and Swenson 2012). Studies taking a variable-centered approach often suggest that adolescents generally tend to view family less positively than their parents, when compared at mean levels (Leung et al. 2016; Ohannessian et al. 2016). Despite the informative findings from this approach, it overlooks the heterogeneity of parent–adolescent dyads. Different parent–adolescent dyads may vary in the direction and magnitude of their discrepancies. For example, in many families, parents and adolescents may have little discrepancies in perceptions of parenting; in some families, parents may perceive parenting more positively than adolescents to a considerable extent; whereas in other families, adolescents may perceive parenting more positively than parents to a notable extent. Whereas certain combinations of direction and magnitude may be adaptive, other combinations may be maladaptive for adolescent development (De Los Reyes et al. 2010; Rote and Smetana 2016).

A person-centered approach is able to empirically explore whether there are distinct groups of parent-adolescent dyads characterized by different combinations of magnitude and direction of parent–adolescent discrepancies in reports of parenting. For example, two recent studies have taken such an approach and indeed demonstrated distinct patterns of parent–adolescent discrepancies in parenting (i.e., maternal monitoring, right to know, knowledge, and positive interaction; De Los Reyes et al. 2010; Rote and Smetana 2016). They both identified three distinct groups for mother-adolescent discrepancies in parenting: no discrepancies, mothers overreported than adolescents, and mothers underreported than adolescents. They also demonstrated that, compared to the no discrepancies group, adolescents exhibited more delinquent behaviors (e.g., truancy) when their mothers overreported (but not when their mothers underreported) than them on parenting These two pioneering studies indicate the utility and importance of taking a person-centered approach to investigate parent–adolescent discrepancies in parenting. Nevertheless, they did not address several other limitations in prior studies, that is, there is still a dearth of studies on configurations of discrepancies in multiple dimensions of parenting, the implications of parenting discrepancies on adolescent academic and physical health domains, and parenting discrepancies in father–adolescent dyads and low-socioeconomic status Mexican immigrant families, which are described below.

Most prior studies on parenting discrepancies examined how discrepancies in each parenting dimension relate to adolescent outcomes separately (Abar et al. 2015; Gaylord et al. 2003; Reidler and Swenson 2012). Although De Los Reyes et al. (2010) and Rote and Smetana (2016) examined combinations of parent–adolescent discrepancies across multiple parenting constructs, the parenting constructs were actually subscales of one larger parenting dimension (e.g., child disclosure, parental knowledge, and parental solicitation, all aspects of parental monitoring). However, the parenting literature has well-recognized the importance of considering multiple dimensions simultaneously to gain a holistic view of parenting styles and how they relate to adolescent outcomes (Darling and Steinberg 1993; Maccoby and Martin 1983; White et al. 2013). Similarly, there may be different combinations of parent–adolescent discrepancies across multiple parenting dimensions, and different discrepancy combinations may have different implications for adolescent outcomes. Thus, the current study extends prior studies to take a person-centered approach to explore parent-adolescent discrepancy patterns simultaneously considering three important dimensions of parenting: parental warmth, control (i.e., monitoring), and inductive reasoning. Parental warmth and control are key distinct dimensions of parenting that have been widely used to derive parenting styles across various populations (Darling and Steinberg 1993; Maccoby and Martin 1983; White et al. 2013). Inductive reasoning is a key element of effective parental communication with their children and is also considered as an important dimension of parenting, particularly in ethnic minority families (Le et al. 2008; Kim et al. 2013; Taylor et al. 2012).

Extant studies on the relation between parent–adolescent discrepancies in parenting and adolescent outcomes primarily focus on either behavioral (e.g., delinquent or risk behaviors; Abar et al. 2015; Córdova et al. 2014; Córdova et al. 2016; De Los Reyes et al. 2010) or psychological domains of adolescent outcomes (e.g., depressive symptoms; Leung et al. 2016; Nelemans et al. 2016; Reidler and Swenson 2012). It seems that studies focused on behavioral domains tend to show more consistent associations between discrepancies and adolescent problems than do studies examining mental health problems. It is possible that parent–adolescent discrepancies in parenting differentially related to distinct domains of adolescent outcomes. However, few, if any, studies have examined the potential implications of parent–adolescent discrepancies in parenting on other important developmental domains such as adolescents’ academic and physical health outcomes. Given that parenting and parent-adolescent relationships have been demonstrated to be influential for adolescent academic and physical health outcomes (Kim et al. 2013; Repetti et al. 2002), discrepancies in parenting may relate to adolescent academic and physical health outcomes as well. To comprehensively test whether parent–adolescent discrepancies in parenting are adaptive or maladaptive for adolescent development, it is necessary to systematically assess multiple key domains of adolescent outcomes. Thus, the current study moves beyond prior studies to include measures of adolescent outcomes in four key domains: behavioral (delinquent behaviors), psychological (depressive symptoms, anxiety, life meaning, and resilience), academic (school engagement and grades), and physical health outcomes (physical functioning problems and sleep quality).

Furthermore, most prior studies of parent–adolescent discrepancies in parenting focus on only mother-adolescent dyads (De Los Reyes et al. 2010; Leung et al. 2016; Ohannessian et al. 2016; Reidler and Swenson 2012; Rote and Smetana 2016), leaving the discrepancies between father–adolescent dyads understudied. Mothers and fathers often play different childrearing roles in Mexican American families: Mothers often are the primary caregivers who may have more interactions with their children and thus may have a more profound influence on their children’s development relative to fathers (Chuang and Tamis-LeMonda 2009; Dumka et al. 2008). Consequently, mother– and father–adolescent discrepancies may have different patterns and have different implications for adolescent outcomes. Relative to mother–adolescent dyads, there may be a larger portion of father–adolescent dyads who hold discrepant views on parenting, but father–adolescent discrepancies may have less profound influence on adolescent outcomes, given that fathers tend to interact with adolescents less frequently (Chuang and Tamis-LeMonda 2009; Dumka et al. 2008). Thus, our study moves beyond prior studies to explore patterns of discrepancies and their links to adolescent outcomes for not only mother–adolescent dyads but also father–adolescent dyads.

Finally, there is a paucity of informant discrepancies research focusing on ethnic minority groups in the United States (Korelitz and Garber 2016). A few prior studies have suggested that informant discrepancies may be even larger in ethnic minority groups compared to European Americans, probably because ethnic minority parents and adolescents are more likely to have different views about parenting due to their various orientations to the ethnic and mainstream American cultures in which parenting practices are different (Guion et al. 2009; Korelitz and Garber 2016). Recently, scholars have pointed out that the implications of informant discrepancies may vary across cultures and socioeconomic status (SES; Rescorla 2016). The two pioneering studies that have taken a person-centered approach to examine parent-adolescent discrepancy patterns relied on samples that were predominantly middle-class European Americans (Rote and Smetana 2016) or African Americans with diverse socioeconomic status (De Los Reyes et al. 2010). Whether and to what extent their findings can be generalized to low-SES Mexican immigrant families remains to be examined. Low-SES Mexican immigrant families experience more challenges in family functioning, such as economic stress and intergenerational cultural conflicts (Lui 2015), and thus parent–adolescent discrepancies in parenting in these families may be more prevalent and more detrimental relative to other populations.

The Current Study

The current study aimed to provide a comprehensive understanding of parent–adolescent discrepancies in reports of parenting in Mexican immigrant families, addressing three specific research questions. First, are there distinct patterns of parent–adolescent discrepancies in parenting when simultaneously considering parental warmth, monitoring, and inductive reasoning? Taking a person-centered approach, we used latent profile analysis (LPA) to explore patterns of parent-adolescent discrepancies using parent-adolescent standardized difference scores in parental warmth, monitoring, and inductive reasoning as indicators. Consistent with the two prior studies on mother–adolescent parenting discrepancy patterns (De Los Reyes et al. 2010; Rote and Smetana 2016), we expected to find three profiles: one group with adolescents’ ratings similar to parents’, one group reflecting adolescents’ lower ratings than adolescents, and one group reflecting adolescents’ higher ratings than parents. The relative distribution of parent–adolescent dyads across these three profiles, however, were expected to potentially differ from these prior studies given our specific population.

Second, do adolescents in different discrepancy groups demonstrate different behavioral, psychological, academic, and health outcomes one year later? We hypothesized that the group with adolescents’ ratings lower than parents’ would demonstrate worse adolescent outcomes compared to the other two groups, whereas, the group with adolescents’ ratings higher than parents’ ratings would demonstrate outcomes similar to or better than the group with adolescents’ ratings similar to parents’ based on findings in prior studies (De Los Reyes et al. 2010; Rote and Smetana 2016).

Third, do patterns of parenting discrepancies and their associations with adolescent outcomes differ across mother-adolescent and father–adolescent dyads? We proposed that compared to mother–adolescent dyads, lager proportion of father–adolescent dyads may be classified into the discrepant groups but that father–adolescent discrepancies may have less profound associations with adolescent outcomes given fathers’ less frequent interaction with adolescents.

Methods

Participants

The current study used a two-wave longitudinal dataset of Mexican immigrant families in the United States. Participants were 604 Mexican American adolescents (54% female) and 595 of their mothers and 293 of their fathers. The adolescents’ ages ranged from 11 to 15 years old (M = 12.41, SD = .97) at Wave 1. The majority of adolescents (76%) were living with both their mothers (M_age = 38.39, SD = 5.74) and fathers (M_age = 40.82, SD = 6.71), and were born in the United States (75%). Median family income was in the range of $20,001 to $30,000. For both fathers and mothers, the median education level was finished middle school. Most of the fathers (87%) and about half of mothers (46%) were employed at least part-time, and most of the parents’ occupations were unskilled laborer (e.g., construction worker, truck driver, mover, restaurant server).

Procedures

Participants were recruited through public records, school presentations, and community recruitment in and around a metropolitan city in central Texas from 2012 to 2015. Because the larger research project, from which the current data come, focuses on adolescent language brokers of Mexican immigrant families, families qualified to participate if parents were of Mexican origin, with a child in middle school who had the responsibility of translating from English to Spanish for at least one parent. If a family met these qualifications, an acquaintance visit was scheduled to provide the family with comprehensive information about the project and procedures. Family consent (for parents) and assent (for children) were acquired at the acquaintance meeting if the family decided to participate in the project. In the formal interview, bilingual and bicultural interviewers read the questions aloud and entered the participant responses on a laptop computer given that many participants have low educational level and cannot read and write well. Questionnaires were prepared in both English and Spanish (first translated to Spanish and then back-translated to English). Both Spanish and English were presented together on the same questionnaires.

In total, two waves of data (with an interval of approximately one year) were collected following the same procedures. Of the 604 families participating in Wave 1483 (80%) families also participated in Wave 2. Each participating family was compensated $60 at Wave 1 and $90 at Wave 2. Attrition analyses were conducted to compare families who participated in both data collection waves and those who dropped out at Wave 2 on demographic variables and all study variables at Wave 1. We found only one significant difference between groups—families who continued participating had higher levels of maternal education, Cohen’s d = .25 [0.05, 0.64], t(591) = 2.41, p < .05, and paternal education, Cohen’s d = 0.44 [0.12, 0.85], t(291) = 3.13, p < .01.

Measures

Parenting variables

At Wave 1, parenting was assessed through measures adapted from the Iowa Youth and Families Project and the measures have been validated in prior studies (Conger et al. 1995; Kim et al. 2013). Mothers and fathers self-reported their own parenting behaviors, and adolescents reported for maternal and paternal parenting behaviors separately on parallel items. Mean, standardized deviation, and cronbach’s alpha reliability for each parenting measure for each informant were presented in Table 1. The measures demonstrated acceptable reliability across informants (as = .64 to .93).

Table 1.

Descriptive statistics and correlations of parenting variables

	1	2	3	4	5	6	7	8	9	M	SD	a	N
1. PR Warmth	–	.48***	.57***	.18**	.10	.07	−.68***	−.32***	−.38***	5.79	0.93	.79	293
2. PR Monitoring	.43***	–	.56***	.07	.07	.00	− .34***	−.75***	−.42***	4.43	0.70	.64	293
3. PR Reasoning	.55***	.48***	–	.05	.04	.03	− .43***	−.42***	−.74***	3.90	0.90	.87	293
4. AR Warmth	.22***	.15***	.08*	–	.67***	.74***	.60***	.33***	.43***	4.95	1.47	.91	558
5. AR Monitoring	.09*	.16***	.09*	.58***	–	.63***	.36***	.60***	.35***	3.63	1.08	.77	561
6. AR Reasoning	.07	.01	.01	.64***	.55***	–	.46***	.37***	.65***	3.54	1.06	.81	560
7. SDS Warmth	−.63***	−.23***	−.37***	.63***	.38***	.45***	–	.51***	.64***	0.16	1.22	–	293
8. SDS Monitoring	−.26***	−.65***	−.30***	.32***	.65***	.41***	.47***	–	.57***	0.24	1.25	–	293
9. SDS Reasoning	−.34***	−.33***	−.71***	.39***	.32***	.70***	.59***	.50***	–	0.17	1.32	293
M	6.07	4.65	4.27	5.18	4.08	3.82	0.01	0.01	0.01
SD	0.78	0.50	0.70	1.27	0.82	0.93	1.25	1.29	1.40
a	.80	.76	.72	.93	.86	.87	–	–
N	595	595	595	604	604	604	595	595	595

Note: Statistics below the diagonal are for maternal parenting; statistics above the diagonal are for paternal parenting

PR Parent Report, AR Adolescent Report, SDS Standardized Difference Scores

^*p < .05

^**p < .01

^***p < .001

Parental warmth Parental warmth was measured with seven items about this affective dimension of parenting (e.g., “Let your child know that you appreciate him/her, his/her ideas, or the things he/she does?” for parental measure; “(Your mother) Let you know that she appreciates you, your ideas, or the things you do?” for adolescent report of maternal warmth.). The rating scale ranged from 1 (never) to 7 (always). Higher mean scores represent higher warmth.

Parental monitoring Parental monitoring was assessed by three items (e.g., “Do you know who your child is with when he/she is away from home?” for parental measure; “Does your mother know who you are with when you are away from home?” for adolescent report of maternal monitoring.) on a 5-point scale ranging from 1 (never) to 5 (always). Higher mean scores represent higher levels of monitoring.

Parental inductive reasoning Parental inductive reasoning was assessed by four items (e.g., “Do you give reasons (explain) to your child for your decisions?” for parental measure; “Does your mother give you reasons (explanations) for her decisions?” for adolescent report of maternal reasoning.) on a 5-point scale ranging from 1 (never) to 5 (always). Higher mean scores represent higher levels of reasoning.

Adolescent outcome variables

In total, adolescents self-reported on nine outcome measures at Wave 2, which span behavioral (delinquent behaviors), psychological (depressive symptoms, anxiety, life meaning, resilience), academic (grades, school engagement), and physical health domains (physical functioning problems, sleep quality).

Delinquent behaviors Adolescents’ delinquent behaviors were measured with 13 items adapted from the Youth Self-Report (Achenbach and Rescorla 2001), including items such as stealing, running away, and lying. Adolescents reported the extent to which the listed behaviors applied to them during the past six months, on a scale ranging from 0 (not at all true) to 2 (often true or very true). Higher mean scores reflect more delinquent behaviors (α = .79).

Depressive symptoms Depressive symptoms were measured by the widely used 20-item Center for Epidemiologic Studies of Depression Scale (CESD; Radloff 1977). Adolescents self-reported how often during the past week they had experienced depressive symptoms, endorsing items such as “Bothered by things usually not bothered by,” on a scale of 1 (rarely or none of the time) to 4 (most or all of the time). Higher mean scores reflected more depressive symptoms (α = .84).

Anxiety Anxiety was measured by four items adopted from prior studies (Reynolds and Richmond 1997; Spitzer et al. 2006). Adolescent self-reported how often they were bothered by the following problems over the last 2 weeks: (1) feeling nervous, (2) worrying about what is going to happen, (3) trouble relaxing, and (4) becoming easily annoyed or irritable, on a scale of 1 (not at all) to 5 (nearly everyday). Higher mean scores reflected higher levels of anxiety (α = .82).

Life meaning Life meaning was measured using three items from the presence subscale of the meaning in life questionnaire (Steger et al. 2006): “I understand my life’s meaning,” “My life has a clear sense of purpose,” and “I have a good sense of what makes my life meaningful.” These items were selected given their relatively high item-scale correlations and their good face validity (Steger et al. 2006). Adolescents self-reported on a scale of 1 (strongly disagree) to 5 (strongly agree). Higher mean scores reflect a greater sense of life meaning (α = .90).

Resilience Resilience was measured using three items from the Connor-Davidson Resilience Scale (Connor and Davidson 2003), for example, “I tend to recover easily after an illness or hardship.” The three-item scale has been validated in prior research (Kim et al. 2017). Adolescents reported on a scale of 1 (Strongly disagree) to 5 (Strongly agree), with higher scores reflecting a greater sense of resilience (α = .73).

Grades Adolescents identified their average grades at school on a 13-point scale ranging from 1 (F) to 13 (A + ).

School engagement School engagement was assessed using four items developed for the purpose of the study, for example, “I am motivated to get good grades in school.” Adolescents reported on a scale of 1 (Strongly disagree) to 5 (Strongly agree), with higher scores reflecting higher levels of school engagement (α = .73). Confirmatory factor analysis showed excellent model fit of the data to a single factor, χ²(2) = . 68, p = .71, CFI = 1.00, RMSEA = .00 [.00, .07], SRMR = .01. All of the items loaded strongly on the latent factor, λs = .59 to .83, p < .001.

Physical functioning problems Physical functioning problems were assessed by three items adopted from the Physical Functioning subscale of the Pediatric Quality of Life Inventory Version 4.0 (Varni et al. 2001). Adolescent reported how much of a problem has the following been for them during the past month: (1) walking more than one block, (2) running, and (3) participating in sport activities or physical functioning, on a scale of 1 (Never a problem) to 5 (Always a problem). Higher mean scores reflect more physical functioning problems (α = .80).

Sleep quality For sleep quality, adolescents reported on one item “During the past month, how would you rate your sleep quality overall?” from the Pittsburgh Sleep Quality Index (Buysse et al. 1989) on a scale of 1 (poor) to 5 (excellent).

Covariates

A set of demographic variables were included as covariates for adolescent outcomes, including adolescent age, gender, nativity (i.e., whether born in the U.S. or not), and parental education given their associations with adolescent outcomes demonstrated in prior studies (Conger and Donnellan 2007; Kwak 2003; Yip et al. 2008). Parents reported on their highest education level on a scale of 1 (no formal schooling) to 11 (finished graduate degree). In addition, maternal and paternal depressive symptoms were included as covariates for all adolescent outcomes because prior studies have demonstrated a relation between parental depressive symptoms and adolescent well-being (Conger et al. 1995; Parke et al. 2004). Mothers and fathers self-reported their depressive symptoms using the same measure (i.e., CESD) and rating scale as adolescent used (α = .88 and 81 for mothers and fathers, respectively).

Analysis Plan

Data analyses were conducted in three steps. First, standardized difference score was computed by subtracting parent standardized report from adolescent standardized report for each parenting variable. Thus, positive standardized difference score indicates that adolescent standardized score is higher than parent score, and negative standardized difference score indicates that adolescent standardized score is lower than parent score. Standardized difference score was recommended by De Los Reyes and Kazdin (2004) and was most widely used in prior studies (Ehrlich et al. 2011; Guion et al. 2009; Leung and Shek 2014). Our use of standardized difference score is consistent with these studies and two studies using similar person-centered analysis to examine patterns of parent–adolescent discrepancies in parenting (De Los Reyes et al. 2010; Rote and Smetana 2016). Relative to raw difference score, standardized difference score has multiple advantages with properties that are critical when examining associations with informant discrepancies: (a) standardized difference score places both informants’ scores on the same metric with the same variance (z distribution) and thus each informant’s score contributes equally to the difference score; (b) standardized difference score was statistically discernable from the informants’ ratings from which it was created; (c) standardized difference score reflects variations in discrepancy scores that exist beyond those which would be normatively expected given intergenerational differences because it centers informants’ scores relative to the mean of their groups before comparison (De Los Reyes and Kazdin 2004; Rote and Smetana 2016).

Second, latent profile analyses were conducted separately for mother–adolescent and father–adolescent dyads. Mother–adolescent and father–adolescent standardized difference score of the three parenting variables (i.e., warmth, monitoring, and reasoning) were used as indicators in the latent profile analyses. Latent profile analyses were conducted using Mplus 7.31 (Muthén and Muthén 1998–2015). Mplus uses the full information maximum likelihood (FIML) estimation method to handle missing data, which enables full usage of all available data in the model. A series of models were specified (i.e., 1 to 5 profiles). Several model fit indices were used to compare models with varying numbers of profiles in addition to examining whether the profiles appeared substantively and conceptually meaningful and qualitatively unique from other profiles in the model. Specifically, Bayesian information criteria (BIC), sample-size adjusted Bayesian information criteria (ABIC), entropy, and the Lo–Mendell–Rubin likelihood ratio test (LMR) were used. Smaller values on the BIC and ABIC are indicative of a better fitting model, and a significant p value on the LMR indicates that a model with k profiles had better fit to the data than a model with k–1 profiles. Moreover, models with reasonable sample sizes in each profile and higher entropy were also given preference.

Third, after participants were classified into different profiles based on their most likely latent class membership according to LPA results, we examined whether later adolescent outcomes differed across profiles using multivariate analysis of covariance (MANCOVA). Two MANCOVA models were analyzed separately for mother–adolescent dyads and father–adolescent dyads. In each MANCOVA model, the dependent variables were the adolescent outcome variables; the independent variable was mother–adolescent or father–adolescent parenting discrepancy profiles. The covariates included adolescent age, sex, nativity, and maternal or paternal depressive symptoms and education levels.

Results

Descriptive Statistics

Table 1 displays descriptive statistics and bivariate correlations among parenting variables (i.e., warmth, monitoring, and reasoning). The correlations between parents’ and adolescents’ reports of parenting variables were generally small (rs = .00 to .22), suggesting substantial parent–adolescent discrepancies in their perceptions of parenting. The correlations between standardized difference scores of warmth, monitoring, and reasoning were moderate to large (rs = .47 to .64), suggesting considerable consistency of discrepancies across the three dimensions of parenting.

Latent Profile Modeling of Parenting Discrepancies

Model fit indices of latent profile analyses are presented in Table 2. Based on the model fit indices and the identification of conceptually meaningful and interpretable profiles, the 3-profile solution and the 2-profile solution were identified as optimal solution for mother–adolescent discrepancies and father–adolescent discrepancies, respectively. Specifically, for mother–adolescent discrepancies, the 3-profile solution and the 4-profile solution seemed to be reasonable solutions with ABIC and BIC values close to the 5-profile solution, and an LMR p-value smaller than 0.05. Nevertheless, we elected the 3-profile solution as the optimal solution because (a) the pattern of the three groups are consistent with prior studies exploring mother–adolescent discrepancies (De Los Reyes et al. 2010; Rote and Smetana 2016), (b) the 3-profile solution had the highest entropy, and (c) the 3-profile solution had reasonable sample sizes in each profile whereas the 4-profile solution had one very small group (n = 19). For father–adolescent discrepancies, the 2-profile and 3-profile solution seemed to be reasonable with ABIC and BIC values very similar to the 4-profile solution. Nevertheless, the 2-profile solution was identified as the optimal solution because it had an LMR p-value smaller than 0.05 and reasonable sample sizes in each profile, whereas the 3-profile solution had an LMR p-value greater than 0.05 and one very small group (n = 12).

Table 2.

Model fit indices for latent profile analysis of parenting discrepancies

	Log-likelihood	N of parameters	AIC	BIC	ABIC	Entropy	p of LMR adj. LRT	Distributiona
Mother–adolescent discrepancies in parenting
1 profile	−3018.544	6	6049.089	6075.420	6056.372	n/a	n/a	595
2 profiles	−2881.684	10	5783.367	5827.253	5795.506	0.584	0.061	318-277
3 profiles	−2803.832	14	5635.664	5697.104	5652.658	0.763	0.000	73-114-408
4 profiles	−2786.326	18	5608.651	5687.645	5630.501	0.754	0.019	71-19-152-353
5 profiles	−2772.783	22	5589.566	5686.115	5616.271	0.753	0.422	19-86-134-17-339
Father–adolescent discrepancies in parenting
1 profile	−1450.620	6	2913.240	2935.321	2916.294	n/a	n/a	293
2 profiles	−1342.708	10	2705.416	2742.218	2710.505	0.745	0.000	112-181
3 profiles	−1313.929	14	2655.857	2707.380	2662.982	0.795	0.077	12-136-145
4 profiles	−1301.049	18	2638.098	2704.341	2647.258	0.734	0.073	93-142-11-47
5 profiles	−1293.137	22	2630.273	2711.237	2641.470	0.775	0.081	1-10-95-138-49

The optimal solution is bolded

^aNumber of dyads being classified into each class

The mean standardized difference scores of parenting variables in each profile were depicted in Fig. 1. For the 3-profile solution of mother–adolescent discrepancies (upper portion of Fig. 1), in the largest group, mothers’ and adolescents’ standardized scores for each of the parenting domains showed minor discrepancies, with difference scores close to zero (labeled “Adolescent Scores Similar to Mother;” n = 408, 69% of the sample). In the second largest group, adolescents reported consistently lower standardized scores in the three parenting domains than mothers (labeled “Adolescent Scores Lower than Mother;” n = 114, 19%). In the third group, adolescents reported consistently higher standardized scores in the three parenting domains than mothers (labeled “Adolescent Scores Higher than Mother;” n = 73, 12%). For the 2-profile solution of father–adolescent discrepancies (lower portion of Fig. 1), the majority of adolescents reported consistently lower standardized scores in the three parenting domains than fathers, with mean difference scores around −.50 standardized deviation (labeled “Adolescent Scores Moderately Lower than Father;” n = 181, 62% of the sample). In the other group, adolescents reported consistently higher standardized scores in the three parenting domains than fathers with mean difference scores between 1 to 1.5 standardized deviation (labeled “Adolescent Scores Higher than Father;” n = 112, 38%).

Fig. 1.

Discrepancy patterns of positive parenting among mother–adolescent (above) and father–adolescent dyads (below)

Comparing Adolescent Outcomes across Parenting Discrepancy Profiles

For mother–adolescent discrepancy profiles, the multivariate test indicated significant group differences for Wave 2 adolescent well-being, F (18, 910) = 2.56, p < .001. The means and standardized deviations for each well-being indicator for each mother–adolescent discrepancy profile are presented in Table 3 along with the F tests results and effect size of group differences. When we observed significant group differences for a given indicator of adolescent well-being, we further compared the marginal means (i.e., means when accounting for all covariates) of the outcomes for each discrepancy group. The Adolescent Scores Lower than Mother group exhibited higher levels of negative outcomes (delinquent behaviors, depressive symptoms, and anxiety) and lower levels of positive outcomes (life meaning, resilience, school engagement, and sleep quality) compared to the other two groups. Despite the notable discrepancies between mothers and adolescents, the Adolescent Scores Higher than Mother group did not significantly differ from the Adolescent Scores Similar to Mother group in adolescent well-being in general. It is of note that the Adolescent Scores Higher than Mother group had significantly fewer physical functioning problems than the Adolescent Scores Similar to Mother group and the Adolescent Scores Lower than Mother group.

Table 3.

Analysis of covariance contrasting wave 1 mother-adolescent parenting discrepancies profiles on wave 2 adolescent outcomes

	ASHM (N = 55		ASLM (N = 86)		ASSM (N = 330)		F statistic		Effect size of group difference (Cohen’s d [95%CI]
	M	SD	M	SD	M	SD	F(2, 463)	p	ASLM - ASHM	ASLM - ASSM	ASHM - ASSM
Delinquent behaviours	0.24	0.18	0.32	0.22	0.25	0.22	5.00	<.01	0.42 [0.37, 0.47]	0.35 [0.31, 0.38]	−0.04 [−0.08, −0.01]
Depressive symptoms	1.50	0.33	1.72	0.45	1.51	0.37	13.80	<.001	0.54 [0.45, 0.63]	0.53 [0.44, 0.57]	−0.04 [−0.13, 0.00]
Anxiety	1.65	0.61	1.85	0.68	1.69	0.65	3.93	.02	0.31 [0.17, 0.47]	0.25 [0.11, 0.32]	−0.06 [−0.22, 0.01]
Life meaning	3.78	0.71	3.45	0.63	3.75	0.82	7.48	<.001	−0.50 [−0.63, −0.31]	−0.38 [−0.52, −0.30]	0.04 [−0.15, 0.12]
Resilience	3.53	0.59	3.33	0.58	3.60	0.67	7.12	<.001	−0.35 [−0.47, −0.19]	−0.42 [−0.55, −0.35]	−0.11 [−0.27, −0.04]
School engagement	3.80	0.53	3.72	0.55	3.93	0.59	5.29	<.01	−0.27 [−0.38, −0.13]	−0.35 [−0.47, −0.29]	−0.11 [−0.25, −0.04]
Grades	9.73	1.79	9.64	1.84	10.12	1.91	2.62	.07	−0.05 [−0.43, 0.43]	−0.25 [−0.64, −0.05]	−0.21 [−0.68, 0.00]
Physical functioning problems	1.30	0.50	1.55	0.66	1.51	0.73	3.06	.05	0.43 [0.29, 0.56]	0.06 [−0.07, 0.14]	−0.30 [−0.43, −0.22]
Sleep quality	3.24	1.19	2.86	1.03	3.12	1.03	3.42	.03	−0.35 [−0.57, −0.04]	−0.25 [−0.47, −0.14]	0.11 [−0.20, 0.22]

Covariates included adolescent age, gender, nativity,maternal educational level, and depressive symptoms. Significant F statistics are bolded. Effect size of group difference was bolded when the group difference was significant according to follow-up pairwise comparison in MANOVA

ASHM Adolescent Scores Higher than Mother, ASLM Adolescent Scores Lower than Mother, ASSM Adolescent Scores Similar to Mother, N represents the number of valid cases in each group for the MACOVA analysis

For father–adolescent discrepancy profiles, the multivariate test indicated significant group differences on Wave 2 adolescent well-being indicators, F (9, 209) = 3.15, p < .001. The means and standardized deviations for each well-being indicator for each father-adolescent discrepancies profile are presented in Table 4 along with the F tests results and effect size of group differences. Compared to Adolescent Scores Moderately Lower than Father group, the Adolescent Scores Higher than Father group exhibited lower levels of delinquent behaviors, depressive symptoms, and anxiety and higher levels of school engagement and grades.

Table 4.

Analysis of covariance contrasting wave 1 father-adolescent parenting discrepancies profiles on wave 2 adolescent outcomes

	ASHF (N = 81)		ASMLF (N = 143)		F statistic		Effect size of group difference (ASHF - ASMLF)
	M	SD	M	SD	F(1, 217)	p	(Cohen’s d [95%CI]
Delinquent behaviors	0.18	0.15	0.28	0.22	14.16	<.001	−0.54 [−0.57, −0.50]
Depressive symptoms	1.45	0.32	1.56	0.37	4.37	.04	−0.32 [−0.38, −0.25]
Anxiety	1.55	0.59	1.79	0.65	7.12	<.01	−0.38 [−0.51, −0.28]
Life meaning	3.70	0.86	3.77	0.75	0.39	.53	−0.09 [−0.28, 0.03]
Resilience	3.55	0.69	3.57	0.66	0.09	.76	−0.02 [−0.17, 0.08]
School engagement	4.09	0.51	3.85	0.55	10.39	<.001	0.43 [0.32, 0.52]
Grades	10.52	1.62	9.97	1.99	6.25	.01	0.29 [−0.06, 0.62]
Physical functioning problems	1.50	0.75	1.43	0.59	.00	.96	0.10 [−0.06, 0.20]
Sleep quality	3.27	1.10	3.08	1.00	1.11	.29	0.18 [−0.05, 0.35]

Covariates included adolescent age, gender, nativity, paternal educational level and depressive symptoms. Significant F statistics are bolded.

ASHF Adolescent Scores Higher than Father, ASMLF Adolescent Scores Moderately Lower than Father, N represents the number of valid cases in each group for the MACOVA analysis

Discussion

Discrepancies between reports from different informants on the same construct are common in clinical and developmental studies of adolescents and pose challenges for researchers and clinicians in regard to how to interpret and address such discrepancies (Achenbach et al. 1987; De Los Reyes et al. 2015; Korelitz and Garber 2016). Extant literature is inconsistent regarding whether parent–adolescent discrepancies in their reports of parenting relate to more positive or more negative adolescent outcomes (De Los Reyes and Ohannessian 2016; Gaylord et al. 2003; Guion et al. 2009; Maurizi et al. 2012). In support of the modified Operations Triad Model (De Los Reyes and Ohannessian 2016), the current study sheds light on this debate by demonstrating distinct patterns of parent–adolescent discrepancies and their different associations with adolescent behavioral, psychological, academic, and physical health outcomes. We found that, in general, adolescents’ reports of lower levels of positive parenting (warmth, monitoring, and reasoning) relative to parents was associated with worse adolescent adjustment; whereas adolescents’ reports of higher levels of positive parenting relative to parents was related to better adolescent adjustment. We also found some variations between mother– and father–adolescent dyads in patterns of discrepancies and how discrepancy patterns relate to adolescent outcomes.

Patterns of Discrepancy

The three profiles we found for mother–adolescent discrepancies are consistent with prior studies, but the proportion of dyads being classified into each profile is somewhat different from prior studies (De Los Reyes et al. 2010; Rote and Smetana 2016). In line with prior studies, the Minor Discrepancies group was the largest group (69%). However, in our sample, more dyads were classified into the Mother Higher group (19%) than in the Adolescent Higher group (12%), whereas in prior studies, there were fewer dyads in the Mother Higher group than in the Adolescent Higher group for discrepancies in similar parenting constructs, including parent–adolescent positive interactions (6.5% vs. 30.2%; Rote and Smetana 2016) and parental monitoring (15% vs. 24%; De Los Reyes et al. 2010). This inconsistency is likely due to the different characteristics of the study samples: whereas the two prior studies focused on samples of primarily European American (Rote and Smetana 2016) or African American families (De Los Reyes et al. 2010), our study used a sample of Mexican immigrant families. Mexican immigrant families experience additional family functioning challenges, such as intergenerational acculturation gaps wherein adolescents acculturate to the mainstream American cultures faster than their parents (Lui 2015; Telzer 2010). In these families, parents are more likely to perceive parenting more (rather than less) positively relative to adolescents probably because immigrant parents rear children in ways consistent with their ethnic cultural values of positive parenting, but their parenting practices may be perceived less positively by their adolescent children who endorse more American cultural values. We encourage future studies to examine what factors lead to different patterns of discrepancies in different populations.

For father–adolescent dyads, the majority were classified into the Father Moderately Higher group (62%), with the remainder classified into the Adolescent Higher group (38%). This distribution is quite distinct from the distribution across profiles for the mother–adolescent dyads, which were most commonly characterized by minor discrepancies (69% of all mother–adolescent dyads). These results are consistent with our hypothesis that father–adolescent dyads would be more likely to have discrepant views of parenting behaviors relative to mother–adolescent dyads. This may be a result of fathers’ and mothers’ differential family roles, such that compared to mothers who are often the primary caregivers, fathers tend to spend less time and have less communication with their children (Chuang and Tamis-LeMonda 2009; Dumka et al. 2008). With less communication, father–adolescent dyads were more likely to disagree with each other (Ehrlich et al. 2015). We did, however, have a smaller sample of father–adolescent dyads (N = 293) than mother–adolescent dyads (N = 595). It is possible that future studies with a larger sample of father–adolescent dyads may find a group with minor discrepancies.

For both mother– and father–adolescent dyads, discrepancies across the three parenting domains showed consistent patterns in each profile. This suggests that parents who tend to report higher (or lower) standardized scores relative to adolescents in one parenting domain also tend to report higher (or lower) standardized scores in other parenting domains. This result is consistent with a prior study demonstrating that parent–adolescent discrepancies in three distinct parenting constructs (i.e., harsh discipline, inconsistent discipline, and low parental nurturance) could be represented by a single latent variable (Guion et al. 2009). These findings suggest that discrepancies across different facets of parenting may be driven by certain common underlying factors. The modified Operations Triad Model (De Los Reyes and Ohannessian 2016) proposes two hypotheses regarding factors driving discrepancies: on the one hand, discrepancies may be induced by adolescents’ increasing autonomy, and thus are normative or beneficial for adolescent development; on the other hand, discrepancies may be due to family functioning problems and thus are harmful for adolescent development. In the following section we discuss the circumstances under which each of these hypotheses may be true.

Discrepancies and Subsequent Adolescent Well-being

In the current study, different patterns of discrepancies had different implications for adolescent well-being. Specifically, for both mother– and father–adolescent dyads, when the parent reported higher (vs. lower or similar) standardized scores than the adolescent across positive parenting domains, the adolescent exhibited higher levels of maladjustment (i.e., delinquent behaviors, depressive symptoms, anxiety) and lower levels of positive adjustment (e.g., school engagement, sleep quality). In many prior studies using a variable-centered approach (Córdova et al. 2014; Gaylord et al. 2003; Guion et al. 2009; Reidler and Swenson 2012), it is often unclear whether significant associations between parent–adolescent difference scores and adolescent outcomes were driven by adolescents’ reports of lower or higher levels of parenting relative to parents. Our findings suggest that discrepancies in the direction that parents perceived parenting more positively than adolescents seem maladaptive for adolescent development, whereas discrepancies in the direction that adolescents viewed parenting more positively than their parents seem adaptive for adolescent development. When parents perceive their parenting practices more positively than adolescents, parents are less likely to be aware of weaknesses in their parenting behaviors and adjust their parenting behaviors in responding to adolescents’ needs (Guion et al. 2009). This can make parent–adolescent communication more difficult and improvements in parenting less likely, which may contribute to adolescent maladjustment. In contrast, adolescents’ perception of higher levels of positive parenting relative to parents may reflect that parenting behaviors fit adolescents’ increasing need for autonomy and it may also indicate adolescents’ understanding of their parents, which may contribute to adolescents’ positive adjustment (Guion et al. 2009; Leung et al. 2016).

Our findings are in line with the two previous person-centered approach studies which also demonstrated that adolescents tend to exhibit more problem behaviors when mothers reported higher (vs. similar) standardized scores than adolescents (De Los Reyes et al. 2010; Rote and Smetana 2016). However, our study extended these two studies and most prior studies in three ways. First, by including multiple domains of adolescent development, our results indicate that discrepancy patterns have significant and consistent implications on adolescent outcomes across domains, thus providing more robust evidence for the utility of parent–adolescent discrepancies in predicting adolescent outcomes. Second, we found that discrepancy patterns in both mother– and father–adolescent dyads can significantly relate to adolescent outcomes, however, the associations appear to vary across measures of adolescent outcomes. Only profiles of mother–adolescent discrepancies significantly relate to physical health measures (i.e., physical functioning problems and sleep quality) and positive measures in psychological domain (i.e., life meaning and resilience). These findings are in line with a prior study which observed that mother–child (but not father–child) discrepancies in parental responsiveness related to lower levels of adolescents’ psychological competence (Leung and Shek 2014). It seems that mothers, often as the primarily caregivers who also tend to be the warmth and support providers, play a more important role in taking care of their children’s physical health and promoting their children’s positive psychological development (Chuang and Tamis-LeMonda 2009; Dumka et al. 2008).

Furthermore, moving beyond prior studies that focused on samples of predominantly middle-class European Americans (Abar et al. 2015; Ohannessian et al. 2016; Rote and Smetana 2016), the current study used a sample of low-socioeconomic status (SES) Mexican immigrant families, an underrepresented population in informant discrepancies literature. It has been shown that children of ethnic minority (especially Mexican immigrant, vs. European American) families and low-SES (vs. high-SES) families have more developmental problems (Conger and Donnellan 2007; Gregory et al. 2010; National Center for Health Statistics 2016). It is important to identify risk and protective factors for child development in low-SES Mexican immigrant families to help close such race-ethnicity and socioeconomic status gaps. The current study indicate that one of the risk factors in these families may be that parents perceive their parenting practices more positively than their adolescent children to a considerable extent. This pattern of parent–adolescent discrepancies may reflect family functioning problems that particularly associate with unique challenges experienced by these families such as intergenerational cultural conflicts and economic stress (Lui 2015). In contrast, adolescents’ more positive perception of parenting than parents may be a protective factor in these families. Interventions aimed at promoting adolescents’ positive perceptions of parenting toward them may benefit their development.

Limitations and Future Directions

Several limitations and caveats should be kept in mind when considering the current study’s implications. First, we used standardized difference scores to represent parent–adolescent discrepancies. Although this is the most widely adopted approach to date, scholars have raised some concerns about this approach (De Los Reyes and Ohannessian 2016; Laird and De Los Reyes 2013). However, our study took a person-centered approach to use standardized difference scores as indicators of latent profile analysis and used profile membership to predict adolescent outcomes, rather than the standardized difference scores themselves. This may circumvent some of the concerns regarding using standardized difference scores as predictors of adolescent well-being (Rote and Smetana 2016). That said, future studies could employ alternative methods (e.g., polynomial regression; Laird and De Los Reyes 2013) to study parent–adolescent discrepancies in parenting, particularly in ethnic minority families, to provide additional knowledge of parenting discrepancies in these families. Second, this is a correlational study, and thus we cannot ascertain causal relationships. Third, although we have a relatively large sample, our sample of fathers is substantially smaller than the sample of mothers because fathers were less likely to participate possible due to lower interest and less available time to participate. Whether similar profiles as found in mother–adolescent dyads also exist in father–adolescent dyads should be further tested in future studies. Finally, our study focused on a relatively large homogeneous sample of low-SES Mexican immigrant families. The results provide important information regarding parent–adolescent discrepancies in parenting in this understudied population and thus can enrich prior literature on parent–adolescent discrepancies. That said, future studies should include samples that are more diverse in ethnicity or SES to directly examine potential moderating effects of ethnicity and SES on the focal relations.

Conclusion

The current study has significant contributions and implications for research involving parents’ and adolescents’ reports of parenting and for research on informant discrepancy by taking a person-centered approach to identify patterns of parent–adolescent discrepancy and how they predict various adolescent outcomes in Mexican immigrant families. First, moving beyond prior informant discrepancy research which mainly focused on identifying sources of informant discrepancies in reports of child mental health problems (Achenbach et al. 1987; De Los Reyes et al. 2015), the current study join recent studies on parent–adolescent discrepancies in parenting (Abar et al. 2015; De Los Reyes and Ohannessian 2016; Guion et al. 2009) to suggest that parent–adolescent discrepancies can provide important information about family processes which are useful in predicting adolescent outcomes. Second, our findings highlight the importance for researchers and clinicians to consider both the magnitude and direction of discrepancies for more accurate understanding of discrepancies. Clinicians may need to pay particular attention to families wherein parents perceive their parenting more positively than their children, as these families may have higher risk for adolescent developmental problems. Moreover, given that father– and mother–adolescent discrepancies demonstrated different patterns and implications for adolescent well-being, researchers and practitioners should consider both dyads to gain a more comprehensive understanding of parenting discrepancies.

Biography

Yang Hou is a doctoral student in the Department of Human Development and Family Sciences at the University of Texas at Austin. Her research interests focus on how socio-cultural factors relate to family processes and adolescent socio-emotional, behavioral, academic, and health outcomes, particularly among minority and immigrant families.

Su Yeong Kim is an Associate Professor in the Department of Human Development and Family Sciences at the University of Texas at Austin. She received her Ph.D. in Human Development from the University of California, Davis. Her research interests include the role of cultural and family contexts that shape the development of adolescents in immigrant and minority families in the U.S.

Aprile D. Benner is an Associate Professor in the Department of Human Development and Family Sciences at the University of Texas at Austin. Her substantive research interests center on the development of low-income and race/ethnic minority youth, investigating how social contexts influence school transitions, experiences of marginalization and discrimination, and developmental outcomes during adolescence.